In this age of information technology, data is inextricably linked to everything. Adding a data analysis component to a company’s strategy seems to be a necessary step in every industry. From the beginning of e-commerce to restaurants and supermarkets, data in various forms helps stakeholders to better understand their business. Today I want to share how my UC Davis practicum partner, REEF Technology, is currently using analytics to help them improve their company’s profits.
As a startup, REEF has a very interesting business concept. What they started doing and what they have done most successfully is their parking business. All over the world, REEF owns its parking lots and has a good track record. However, as their parking business matured, REEF had a very novel idea:
“According to marketing materials, Reef creates ‘thriving hubs for the on-demand economy’ by ‘reimagining the common parking lot’. By bringing in utilities like electricity, gas, and water, and setting up ‘proprietary containers’, the company hopes to turn parking lots into reconfigurable community hubs. ” — Our Ghost-Kitchen Future, Anna Wiener
In short, it means that REEF used the rich space and physical conditions of the parking lot to set up a mobile kitchen and start a take-out service. What seems to be a unique condition for them is that they can save the rental fees of the house. Soon, REEF had over two hundred mobile kitchens in North America, and the business also seemed to have a very bright future.
But as time went on, REEF gradually found that they faced a problem with labor distribution. The main reason is that the demand for orders and sales varies from season to season, from time to time, from city to city, and even from street to street, which means that the required labor force is also very different. So the question arises, if the same workforce is uniformly assigned to each kitchen, two situations will arise, either there will be a shortage of workforce, which will not be able to fulfill the share of sales, or there will be a surplus of the workforce, which will have a large number of unnecessary labor expenses. So REEF decided to use the analytics team to analyze the data, cutting expenses and increasing profits through optimal modeling. Our student team is the core component of this project.
What we needed to deliver was a dashboard for REEF’s internal kitchen managers, allowing them to forecast the daily sales budget and the corresponding budgeted workforce at the beginning of each week with a simple keystroke, helping the managers to make a workforce allocation strategy for the week as soon as possible. There are many complex processes involved, such as how to tableau to a database and call data in Python at the same time, and deciding what model to use for the most optimal prediction. Although the project has not been delivered yet, we are enjoying the learning process very much.
The project looks very impressive, but the risks and benefits are actually indistinguishable. If this program is tested and ends up having good results, then we can at least help REEF reduce labor payments by more than 20% at this stage, and without affecting marketing revenue. But everything related to prediction carries the risk that the results may not be accurate, and if we make a mistake in our prediction, or if there is a bug in the middle of the calculation, then the final product may not only fail to help REEF solve the problem, but may become very misleading. But that’s the data, isn’t it? If you use it wisely, then it will bring you unexpected gains, but if you make a mistake, then you may also pay the corresponding price.
